INDEX

Explanations

the concept of comparison or evaluating differences and similarities

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

peria

-0.77

FTWARE

-0.72

 WARRANT

-0.63

gat

-0.57

KING

-0.56

mare

-0.56

roup

-0.56

LEASE

-0.55

 Related

-0.55

estern

-0.55

POSITIVE LOGITS

 frankly

0.81

pires

0.80

 fortunately

0.77

 rightly

0.73

fortunately

0.71

 understandably

0.70

 alas

0.69

 somew

0.69

 admittedly

0.67

 unfortunately

0.67

Activations Density 1.383%

the concept of comparison or evaluating differences and similarities

No Comments

No Known Activations